Pay for performance initiatives reward providers for clinical care that meets evidentiary standards for high quality. By definition such initiatives cannot cover clinical interventions and procedures of uncertain effectiveness, and unfortunately, the effectiveness of much of what we do in clinical practice is uncertain. Of greatest concern, given relentlessly escalating healthcare costs and potential harm to patients, are invasive therapeutic procedures. While most Americans understand the role of the Food and Drug Administration in the evaluation, approval, and regulation of pharmaceuticals, few realize that surgical procedures find their way into widespread use without undergoing rigorous testing like new drugs must. To the extent that inefficacious and/or harmful procedures find their way into widespread use in the community, health care resources are wasted and individuals are exposed to unnecessary risks. We have conceptualized a variant of pay for performance that would enable these quality improvements. [unreadable] [unreadable] We call it "pay for evidence." It is based on the principle that when evidence is lacking about the best clinical course of action, then the best clinical performance is one that generates evidence. Such evidence would become the basis for future individual and organizational health care decision-making. We are applying for support to hold a national invitational conference involving approximately 80 experts in spring of 2008 to develop a pay for evidence approach to address the problem of the widespread dissemination of invasive therapeutic procedures before they have undergone rigorous testing of their benefits, harms, and alternatives. The purpose of the conference is to develop a research agenda for studying a pay for evidence approach. The proposed conference addresses the "Payment and Organization" priority area for large conferences sponsored by AHRQ. Objectives are (1) to inform about the problem of widespread dissemination of invasive therapeutic procedures in the US before they have undergone rigorous testing of their benefits, harms, and alternatives; (2) to begin to understand the concerns and conflicts of stakeholders about a pay for evidence program; (3) to discuss program design aspects, including the configuration of financial incentives; and (4) to discuss the governance, oversight, and operational structures that would be needed to support a pay for evidence program. Objectives will be accomplished by means of pre-conference background papers and six concurrent work groups at the conference. Products, to be disseminated via a website and will include reports from the six work groups, a conflict assessment paper summarizing the interests of stakeholders in a pay for evidence program and their concerns, conflicts, and common ground, and a paper describing a pay for evidence national research agenda. [unreadable] [unreadable] [unreadable]